Вы находитесь на странице: 1из 5

Open Access

Original Article

Otitis media with effusion: Accuracy of tympanometry in detecting

fluid in the middle ears of children at myringotomies
Khurshid Anwar1, Saeed Khan2, Habib ur Rehman3,
Mohammad Javaid4, Isteraj Shahabi5
Objective: (1) The diagnostic accuracy of tympanometry in detecting fluid in the middle ear space in
children with otitis media with effusion by comparing its findings with those of myringotomies. (2) Identify
the age group most commonly affected by OME.
Methods: This prospective study was conducted at the Department of ENT& Head and Neck Surgery,
Postgraduate Medical Institute Hayatabad Medical complex, Peshawar from July 1, 2012 to April 30,
2015. Patients with suspicion of OME underwent tympanometry and later myringotomies. Using Jerger’s
classification, Type B tympanogram with normal canal volume was considered as conclusive evidence of
fluid in the middle ear space. Its findings were compared with those of the respective myringotomies. From
the data collected, the accuracy, sensitivity, specificity, positive predictive value and negative predictive
values were calculated.
Results: A total 117 ears of 63 patients were operated. The age range was 3 to 12 years. The commonest
age group (58.7%) affected by OME was 6-8 years. Type B tympanogram with flat curve and normal canal
volume was obtained in 71.4% of the ears. Comparison with myringotomy findings showed TP 85, TN 13,
FP 5 and FN 14. The diagnostic value of tympanometry was; Sensitivity 85.85%, Specificity 72.22%, PPV
94.44%, NPV 48.14% and Accuracy of 83.76%. P value calculated using chi square test showed that there was
significant difference between tympanometry and myringotomy findings in OME (p < 0.05).
Conclusions: OME is common in age group 6-8 years. Tympanogram Type B with normal canal volume is
fairly sensitive in diagnosing this condition. However for occurrence of false positive results, final decision
regarding management should be made on clinical findings and other supportive audiological tests.
KEY WORDS: Accuracy, Hearing impairment, Otitis Media with Effusion, Tympanometry.
doi: http://dx.doi.org/10.12669/pjms.322.9009
How to cite this:
Anwar K, Khan S, Habib-ur-Rehman, Javaid M, Shahabi I. Otitis media with effusion: Accuracy of tympanometry in detecting fluid in
the middle ears of children at myringotomies. Pak J Med Sci. 2016;32(2):466-470.
doi: http://dx.doi.org/10.12669/pjms.322.9009
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

INTRODUCTION infection.1 Symptoms usually involve hearing loss

or aural fullness but typically do not involve pain or
Otitis media with effusion (OME) is defined as
fever. The condition is said to be chronic when the
fluid in the middle ear and, sometimes, the mastoid
fluid accumulation persists beyond 12 weeks.2
air cell system without signs or symptoms of ear
OME occurs commonly during childhood, with as
Correspondence: many as 90 percent of children (80% of individual
Dr. Khurshid Anwar,
ears) having at least one episode of OME by age
Senior Registrar, ENT Department, 10.3 Unlike acute otitis media, the prevalence of
PGMI, Hayatabad Medical Complex,
Peshawar, Pakistan.
chronic otitis media with effusion is unknown.
E-mail: dr.khurshidanwar@yahoo.com Several studies have reported various estimates
* Received for Publication: September 18, 2015
of the condition according to age. Rates vary from
* Revision Received: January 27, 2016 13% at age one year, 14% at age two years, 10% at
* Revision Accepted: January 31, 2016 age three years and 2.8% among children aged 7-8

466 Pak J Med Sci 2016 Vol. 32 No. 2 www.pjms.com.pk

Otitis media with effusion

years.4-6 The resultant hearing impairment has its Inclusion Criteria:

own sequelae. However its long-term impact on 1. Patients between the age of 3-12 years and
child developmental outcomes such as speech, belonging to both the sexes.
language, intelligence, and hearing remains unclear. 2. All patients undergoing tympanometry
Children with Eustachian tube dysfunction, followed by myringotomies.
adenoid hyperplasia, nasal allergy, cleft palate, 3. Patients with types A, B and C tympanograms
Down syndrome and other craniofacial anomalies with clinical evidence of OME.
are at high risk for developing OME.7 Recently 4. Pure tone audiometry showing conductive
GERD has also been implicated in OME in young hearing loss with A-B gap of >30 dB in better
children.8 Although rare, OME also occurs in adults. ear at first visit.
This usually occurs following upper respiratory Exclusion Criteria:
infection, severe nasal allergies and rapid air 1. Ears with otoscopic evidence of
pressure changes during flight or scuba diving. The tympanosclerosis.
incidence of prolonged OME in adults can occur but 2. Patients with frankly discharging ears or having
is rare and is much less common than in children.9 evidence of cholesteatoma.
Correct diagnosis is vital for the management of 3. Patients with Type B tympanograms with flat
children with OME. The clinical diagnosis of OME curve and above normal canal volume.
is made by history, otoscopy, pneumatic otoscopy 4. OME persisting for less than 3 months.
and impedance audiometry. The otoscopic findings Data Collection Procedure: Patients fulfilling the
in OME are mainly different combinations of laid down criteria were included in the study. The
retraction of the pars tensa and wide variations procedure was explained and Informed consent
colour of the tympanic membrane. Tympanometry obtained from the parents. Ethical approval for
provides useful quantitative information about the the study was obtained from the institutional
presence of fluid in the middle ear, mobility of the ethical committee. A detailed history was obtained
middle ear system, and ear canal volume. Its use regarding hearing impairment, its duration & mode
has been recommended in conjunction with more of onset, progression of symptoms and performance
qualitative information (e.g. history, appearance at school. Further enquiry was made to look for the
and mobility of the tympanic membrane) in the presence or otherwise of nasal allergies, recurrent
evaluation of otitis media with effusion. episodes of upper respiratory tract infections,
A type B tympanogram with flat curve and normal cleft palate repair, snoring & sleep disturbances
canal volume is considered diagnostic of OME. and concomitant systemic disorders. A detailed
Compared with all other types of tympanograms it ENT and systemic examination including indirect
has a sensitivity of between 56 and 73 percent and a nasopharyngoscopy were undertaken. Otoscopic
specificity of between 50 and 98 percent in detecting examination was focused to look for signs of OME.
OME confirmed surgically.10 The aim of our study Abnormal coloration and retraction of tympanic
was to assess the accuracy of this particular type of membrane, air bubbles or fluid level in the middle
tympanogram in OME. ear cavity were taken as positive signs of OME.
Whisper test and conversational voice test, where
METHODS appropriate, were used to estimate roughly the
This prospective study was conducted at the hearing impairment. Tuning fork tests such as
Department of ENT& Head and Neck Surgery, Rinne, Weber and Absolute Bone Conduction tests
Postgraduate Medical Institute Hayatabad Medical carried out to determine the type of hearing loss.
complex, Peshawar from July 1, 2012 to April 30, Patients with clinical suspicion of OME underwent
2015. It was a prospective and comparative study tympanometry. The tympanograms were obtained
using the non-probability convenience sampling from different centers but all using 226 Hz probe
technique. tone. Normal ear canal volume was taken as 0.3 - 1
Sample Size: The study included 63 cases with mL and the curve was considered flat when it had
tympanometric evidence of OME undergoing no discernible peak over a pressure range of +200
myringotomies during this period. Sample size was daPa to -400 daPa. Pure Tone Audiogram (PTA)
calculated using 13% proportion of Otitis media was also obtained in selected children. X-ray films
with effusion, 95% confidence level and 5% margin of the nasopharynx for adenoids were obtained in
of error using WHO software for sample size cases suspected of having adenoid hypertrophy.
determination. The tympanograms were classified using Jerger’s

Pak J Med Sci 2016 Vol. 32 No. 2 www.pjms.com.pk 467

Khurshid Anwar et al.

Table-I: Age groups and duration of symptoms. Table-II: Tympanogram Types.

Frequency Percent Cumulative Frequency Percent Cumulative
Percent Percent
Age 3-5 13 20.6 20.6 Left Type A 3 4.8 4.8
6-8 37 58.7 79.4 Tympa- Type B 47 74.6 79.4
9-12 13 20.6 100.0 nogram Type C 13 20.6 100.0
Total 63 100.0 Types Total 63 100.0
Duration of 3-6 months 17 27.0 27.0 Right Type A 8 12.7 12.7
Symptoms 6-12 months 30 47.6 74.6 Tympa- Type B 43 68.3 81.0
> 1 year 16 25.4 100.0 nogram Type C 12 19.0 100.0
Total 63 100.0 Types Total 63 100.0

classification as: 1. Type A 2. Type B (flat curve and FP), PPV = (TPx100) / (TP+ FP) and NPV= (TNx100)
normal canal volume) and 3. Type C. Patients with / (TN+ FN).
suspected OME were booked for myringotomies Statistical Analysis: The data was recorded on
and any other concomitant surgery such as a proforma and the descriptive statistics were
adenoidectomy. Haematological and other relevant analyzed using SPSS 16 for Windows to determine
investigations were carried out to determine frequencies for variables like gender, age,
patients’ suitability for surgery and fitness for duration of symptoms, types of tympanograms
general anaesthesia. All myringotomies were and myringotomy findings. Chi- square test was
carried out through a radial incision in the antero- applied to determine the significance of findings at
inferior quadrant using a general inhalational tympanometry and myringotomies.
anaesthetic agent. Type B tympanogram with flat
curve and normal canal volume alone was taken RESULTS
as conclusive evidence for the presence of fluid The study included 43 males and 20 females
in the middle ear space. The operative findings at and a total 117 ears of 63 patients were operated.
myringotomy were recorded and thus labeled as: The M: F ratio was 2.15:1. The age range was 3
I. True Positive (TP) when fluid was present and to 12 years with mean age of seven years and a
II. False Positive (FP) when no fluid was aspirated. standard deviation of +/-2.124. The commonest
In cases where the tympanograms were either age group (58.7%) affected by OME was 6-8 years.
Type A or Type C, the findings were categorized The majority (47.6%) of children undergoing
as: III. True Negative (TN) when no fluid was myringotomy had OME persistent for 6-12 months
aspirated and IV. False Negative (FN) when fluid followed by those (25.4%) who had it for more than
was present. The accuracy, sensitivity, specificity, one year (Table-I). Type B tympanogram with flat
positive predictive value (PPV) and negative curve and normal canal volume was obtained in
predictive value (NPV) of Type B tympanogram 71.4% of the ears examined. The frequency and
with flat curve and normal canal volume were types of tympanograms obtained in the Left and
calculated using the following formulae: Accuracy Right ears are shown in Table-II. On the right side
= (TP+TN) x100/ (TP+TN+FP+FN), Sensitivity = there were 68.3% TP, 12.7% TN and 11.1% FN as
(TPx100) / (TP+FN), Specificity = (TNx100) / (TN+ shown in Table-III. On the left side there were

Table-III: Comparison of tympanogram and myringotomy findings (Right Ear)

Rt Myringotomy findings Total
Ear not operated Fluid present Fluid absent
Rt Tympanogram Type Type A Count 2 0 6 8
% of Total 3.2% 0.0% 9.5% 12.7%
Type B Count 0 43 0 43
% of Total 0.0% 68.3% 0.0% 68.3%
Type C Count 3 7 2 12
% of Total 4.8% 11.1% 3.2% 19.0%
Total Count 5 50 8 63
% of Total 7.9% 79.4% 12.7% 100.0%
Calculated p-value for the Right Ear=0.000

468 Pak J Med Sci 2016 Vol. 32 No. 2 www.pjms.com.pk

Otitis media with effusion

Table-IV: Comparison of tympanogram and myringotomy findings (Left Ear)

Lt Myringotomy findings Total
Ear not operated Fluid present Fluid absent
Lt Tympanogram Type Type A Count 1 0 2 3
% of Total 1.6% .0% 3.2% 4.8%
Type B Count 0 42 5 47
% of Total .0% 66.7% 7.9% 74.6%
Type C Count 3 7 3 13
% of Total 4.8% 11.1% 4.8% 20.6%
Total Count 4 49 10 63
% of Total 6.3% 77.8% 15.9% 100.0%
Calculated p-value for the Left Ear=0.000

66.7% TP, 8% TN, 8% FP and 11% FN as depicted involving 600 children in different schools. Using
in Table-IV. Out of the total 9 non- operated ears, tympanometry as a screening tool, they found OME
4 occurred on the left side. The tympanometry in 13% of these children. Type B curve was found
and myringotomy findings crosstables for both in 88.5% and type C curve was obtained in 11.5%
the left and right ears and Chi-square test applied of these children.13 These findings are in contrast
showed that with respect to determination of fluid to those of our study. The types of tympanograms
in the middle ear, there was significant difference types obtained in our patients were; Type B (flat
between tympanometry and myringotomy findings curve, normal canal volume) 71.4%, Type C 19.84%
on both the left and right sides (p  <0.05).. The and Type A in 8.7%. Analyzing papers with the
diagnostic value of tympanometry calculated was; findings at myringotomy as the reference ‘gold’
Sensitivity 85.85%, Specificity 72.22%, PPV 94.44%, standard, suggest that a type B tympanogram is the
NPV 48.14% and overall Accuracy of 83.76%. most frequently obtained type in OME, a type A
is infrequently associated with OME and a type C
DISCUSSION falls in between.14 These findings are in agreement
with our current study.
Otitis media with effusion (OME) is a common
Age and climate are well known factors that
but treatable cause of deafness in children. It leads
influence the occurrence of OME. OME is usually
to delay in speech acquisition, behavioral problems
found in the relatively younger 3-5 years age group.
and poor performance at school depending on
In a large study in China involving 2902 children
the age at which it affects the child. There is a
aged 2-8 years, the point prevalence of OME was
need to diagnose it correctly at an earlier stage.
Tympanometry in conjunction with history and 4.3%. By age group, the findings were 14.0% in two
clinical examination is the method most commonly years old, 8.3% in 3 years old, 5.0% in 4 years old,
used. It has been confirmed as sensitive and fairly 4.9% in five years old, 2.8% in 6 years old, 1.7% in 7
specific in identifying children with material hearing years old, and 3.2% in eight year old.15 Our findings
loss associated with OME. The affected children that the 6-8 years age group was most commonly
should be observed closely by serial tympanometry affected are supported by Okur E and colleagues
as some 50% of such cases resolve after three months who in their study involving 2930 children found
and do not justify further management unless the the highest point prevalence of 10.4% in the same
condition recurs. Exceptions to this policy are those age group.16 In a Nepalese study by Mark A and
children having a pure-tone average in the better colleagues found the peak age affected was 10 years
ear of > 30 dB HL at their first visit. In these children (23.1%) in contrast to the findings of this and the
the probability of persistence of OME is greater Chinese study.17
than 80%.11 A Turkish study using confirmation of middle
The data on prevalence of OME is highly varied ear effusion by myringotomy as the gold standard,
in the literature and in our country the literature is found that tympanometry had sensitivity of 96% and
scarce on the subject. A study conducted at the Holy a positive predictive value of 92%. There was a false
Family Hospital by Tallat Najeeb and colleagues positive rate of 8 percent.18 Five false positive cases
using otoscopy and tympanometry found OME in occurred in our study. One possible explanation
7% of the 563 children examined.12 Another study could be the fact that inhalational anaesthetic can
was conducted by Tallat Jabeen and colleagues itself aerate the middle ear giving a ‘false’ dry tap.10
in the twin cities of Rawalpindi and Islamabad A similar study from Mosul, Iraq, using fluid tap

Pak J Med Sci 2016 Vol. 32 No. 2 www.pjms.com.pk 469

Khurshid Anwar et al.

at myringotomy as gold standard and the type B 10. Johansen EC, Lildholdt T, Damsbo N, Eriksen EW.
tympanogram with flat curve as indicative of OME Tympanometry for diagnosis and treatment of otitis media
in general practice. Fam Pract. 2000;17:317-322.
reported an accuracy of 71.4%, sensitivity 97.3%, 11. Haggard MP, Gannon MM, Birkin JA, Bennett KE, Nicholls
specificity 57.2%, positive predictive value 55.3% EE, Browning GG, et al. Risk factors for persistence of
and negative predictive value 97.5%. The Type B bilateral otitis media with effusion. Clinical Otolaryngology.
flat curve was obtained in 62% of these patients.19 2001;26:147-156.
12. Najeeb T, Chohan A, Baig M, Naqi SA. Frequency and
CONCLUSIONS etiology of secretory otitis media and its morbidity in
children. JRMC. 2008;12(2):92-94
OME is more common in age group 6-8 years. 13. Jabeen T, Malik SN, Chattha RU. Frequency of secretory
otitis media in children of age 3 to 5 years in Rawalpindi
Tympanometry with a resultant Type B flat curve and Islamabad. RMJ. 2013;38(1):52-55.
with normal canal volume is fairly sensitive and 14. Orchik OJ, Dunn JW, McNutt L. Tympanometry as a
reliable technique in diagnosing this condition predictor of middle ear effusion. Arch Otolaryngol.
in children aged 3-12 years. However due to 1978;104:4-6.
15. Zhang QI, Wei J, Xu M, Zhang Q, Zhang X, Zhang Z et al.
occurrence of false positive results, final decision Prevalence of otitis media with effusion among children in
regarding management of such children should Xi’an, China: A randomized survey in China’s mainland.
be made on clinical findings and other supportive Ann Otol Rhinol Laryngol. 2011;120(9):617-621.
audio logical tests. 16. Okur EI, Yildirim I, Akif Kilic M, Guzelsoy S. Prevalence of
otitis media with effusion among primary school children in
Declaration of interest: We, the authors, have no Kahramanmaras, in Turkey. Int J Pediatr Otorhinolaryngol.
conflicts of interest to declare in relation to this 17. Mark AI, Matharu V, Dowswell G, Smith M. The point
article. prevalence of otitis media with effusion in secondary school
children in Pokhara, Nepal: a cross-sectional study. Int J
REFERENCES Pediatr Otorhinolaryngol. 2013;77(9):1523-1529.
18. Kemaloglu YK, Beder L, Sener T, Goksu N. Tympanometry
1. American Academy of Family Physicians; American and acoustic reflectometry in ears with chronic retraction
Academy of Otolaryngology-Head and Neck Surgery; without effusion. Int J Pediatr Otorhinolaryngol.
American Academy of Pediatrics Subcommittee on Otitis 2000;55:21-27.
Media With Effusion. Otitis media with effusion. Pediatrics. 19. Mutallib AA. Middle ear effusion: Tympanometry versus
2004;113:1412-1429. operative findings. Ann Coll Med Mosul. 2005;31(1):22-27.
2. Bluestone CD. State of the art: Definitions and classifications.
In: Liu DJ, Bluestone CD, Klien JO, Nelson JD. (eds). Recent Authors’ Contributions:
advances in otitis media with effusion. Proceedings of the 3rd
International Conference. Ontario: Decker and Mosby. 1984. Khurshid Anwar conceived the idea, designed the
3. Shekelle P, Takata G, Chan LS. Diagnosis, history, and study, collected data and wrote and critically reviewed
late effects of otitis media with effusion. Agency for
the manuscript. Saeed Khan did data collection and
healthcare research and quality (AHRQ) Publication No.
03-E023. Evidence Report/Technology Assessment No. 55 statistical analysis. Habib ur Rehman was involved in
(Prepared by Southern California Evidence-based Practice collecting the data and helped in searching the literature.
Center under Contract No 290-97-0001, Task Order No. 4). Mohammad Javaid helped in data collection and editing
Rockville, MD: 2003. the manuscript. Isteraj Shahabi was involved in data
4. Croteau N, Hai V, Pless IB, Infante-Rivard C. Trends in collection, reviewing the manuscript and approved the
medical visits and surgery for otitis media among children. final submission of the study. Khurshid Anwar takes
Am J Dis Child. 1990;114:535-539. the responsibility and is accountable for all aspects of the
5. Mark A, Matharu V, Doswell G, Smith M. The point
work in ensuring that questions related to the accuracy
prevalence of otitis media with effusion in secondary
school children in Pokhara, Nepal: a cross-sectional study. or integrity of any part of the work are appropriately
Int J Paediatr Otorhinolaryngol. 2013;77(9):1523-1529. investigated and resolved.
6. Marchisio P, Principi N, Passali D, Salpietro DC, Boschi G, Authors:
Chetri G et al. Epidemiology and treatment of otitis media
1. Khurshid Anwar,
with effusion in the first year of primary school. Acta
Senior Registrar,
Otolaryngol. 1998;118(4):557-562.
2. Saeed Khan,
7. Alho OP, Oja H, Koivu M, Sorri M. Risk factors for chronic Trainee Medical Officer,
otitis media with effusion in infancy: Each acute otitis media 3. Habib ur Rehman,
episode induces a high but transient risk. Arch Otolaryngol Assistant Professor,
Head Neck Surg. 1995;121:839-843. 4. Mohammad Javaid,
8. Kubba H, Pearson JP, Birchall JP. The aetiology of otitis Associate Professor,
media with effusion: a review. Clinical Otolaryngology. 5. Isteraj Shahabi,
2000;25:181-194. Professor & Head,
9. Finkelstein Y, Ophir D, Talmi YP. Adult-onset otitis 1-5: Department of ENT and Head & Neck Surgery,
media with effusion. Arch Otolaryngol Head Neck Surg. PGMI Hayatabad Medical Complex,
1994;120(5):517-527. Peshawar, Pakistan.

470 Pak J Med Sci 2016 Vol. 32 No. 2 www.pjms.com.pk